The Google I/O 2026 keynote has come and gone, and there's no kidding about Gemini being top priority. While a vast majority of it was dedicated to new Gemini models, Google saved the tail end of the keynote to show off new "intelligent eyewear" devices that are coming out this fall.
In the past, the event has been tailored to show off new tools and features for its other platforms, like Android 17 and Android Auto, but it's interesting how AI continues to be the big focus, especially with Gemini.
Despite being shown off at last week's Android Show, there wasn't any time devoted to other Googlebook laptops that Google previously teased from other hardware makers.
Here's all the biggest announcements from Google I/O 2026 .
Biggest Google I/O 2026 Announcements
- Google Search just got the biggest upgrade in nearly 30 years
- Google unveils Gemini Spark — a '24/7 personal AI agent'
- Samsung's 'Intelligent Eyewear' glasses just launched at Google I/O, and they're coming this fall
- Google just launched Gemini 3.5 Flash — here’s all the upgrades
- Google’s new 'Universal Cart' uses AI to track deals and shop across apps
- Ask YouTube 'entirely reimagines' how you find videos — how it works
- Google’s new $100 AI Ultra plan just changed the AI race
How to watch Google I/O 2026
Google I/O 2026 — Live updates
Google I/O starts soon!
We're just a couple of hours away from the official news-dropping madness getting underway. If you're a fan of Android, Gemini or even if you just casually use Google products like Search, there's sure to be something to get you excited at the event once it kicks off.
Be sure to keep yourself locked to this live blog, as we'll bring you every important piece of information as it happens.
Are you excited for I/O?
Google I/O is about to start and we're excited to see what the company has in store for us. But what are you most excited about?
Hit the poll above and let your voice be heard!
Tom's Guide EIC weighs in
"I'm personally excited to see the new Google and Samsung glasses and how they'll stand out versus Meta Ray-Ban. Hopefully, we'll learn more about unique features and see the designs from Warby Parker and Gentle Monster. Just keep in mind that these glasses won't have a display..." said Tom's Guide Global Editor in Chief Mark Spoonauer.
Gemini Remy could make its debut
Google has kept quiet about what I/O will have to offer, but it has teased a future of agentic AI. This means we may see it unveil some details about "Remy," the codename attached to Google's own personal AI agent.
According to reports, Remy is "your 24/7 personal agent for work, school and daily life, powered by Gemini." It will apparently be able to interact with other Google services and be more advanced than the agentic features Google has already started rolling out to the Gemini app. That means handling tasks proactively, learning your own preferences, and keeping track of things that matter.
Some changes have already come to Gemini app
Gemini is likely to get the lion's share of reveals at I/O 2026, but eagle-eyed users have already spotted some changes rolling out to the Gemini app. There seems to be a new "Thinking level" menu rolling out to the app's model picker, with "standard" and "extended" options added to the Fast and Pro models.
The Gemini app also seems to be prepping more integrations with third-party apps, with support documents suggesting support for Canva, Instacart, and OpenTable is on the way. Which means Gemini should be able to control these apps on your behalf, based on prompts you give it.
No doubt we'll hear more during the I/O conference.
Gemini 3.5 coming to I/O?
A social media post by a Google DeepMind employee seems to confirm that Gemini 3.5 exists. This further bumps expectations that Google will announce the new model at I/O 2026, but we'll have to wait and see.
Pricing has allegedly been posted for the new AI model on X, further making us think it'll launch today.
What about the new Google Home speaker?
While I/O may not be a hardware-focused event, there is still one lingering piece of hardware that hasn't been released — and now would be the perfect time to reveal an official date.
I'm talking, of course, about the Google Home Speaker that was first teased last year. Google said it would be arriving in Spring 2026. Since Spring is currently ongoing, it's about time we got some concrete information. The last thing we need is for this speaker to become the next Galaxy Home.
Timing reminder!
In case you forgot, Google's I/O 2026 keynote kicks off at 10 a.m. PT / 1 p.m. ET / 6 p.m. BST on May 19 (that's today), and you can tune into the event through the YouTube stream embedded above.
Of course, the best thing you can do is keep this live blog open, as not only will we tell you everything as it happens, but we'll explain why you should care about each bit of news.
Googlebook's OS needs a name
Google may have revealed a bunch of details about Googlebooks, including the fact that they will be running on an Android-based operating system, but there's still a bunch we don't know. Like a name, because so far we've just been calling it Googlebook OS.
Rumors suggested that the name AluminumOS could be on the cards, but that might just have been a codename. Now that Googlebooks are confirmed, it's about time Google cleared the rest of those little details up.
Smart glasses may be the exception
While we don't expect to see any mobile hardware at I/O, Google has confirmed we will be seeing more on the smart glasses front. We don't know a lot more than the fact smart glasses will be at I/O in some capacity, but there's a chance we may get our first details about Android XR-powered glasses you can actually buy.
Rumor has it that Samsung's Galaxy Glasses will launch at Galaxy Unpacked in July, and considering Google's close partnership with Samsung, there's a good chance we'll get some news about the AI specs. Though I can't imagine Samsung will allow a full reveal at someone else's event, when it's own is only a couple of months away.
Live stream has started
We're less than an hour out from the start of the Google I/O 2026 keynote. However, up on YouTube, the livestream has already started. We're currently being serenaded by Jellectronica AI music created by Google's Lyria 3. Meanwhile, a video feed shows jellyfish from the Monterey Bay Aquarium who were tracked using Google's AI object detection tech.
We're on the ground live
Our own Jason England and Paul Antill have just arrived at the Shoreline Amphitheatre in Mountain View, California. We'll have more live coverage of the event once they make their way inside to get ready for the Google I/O keynote which is set to start in 30 minutes. So stay tuned!
And we're in
Jason and Paul have now made it inside Google I/O and are getting ready to cover the keynote live from Mountain View. I just spoke with Paul on the phone and he and Jason are currently making their way to their seats. From what I could hear, things sounded pretty busy at the Shoreline Amphitheatre with less than 20 minutes to go before the keynote kicks off.
Game together while you wait
The hosts up on stage at Google I/O are currently playing an interactive, browser-based side scrolling game called Infinite Scaler before the keynote starts. In the game, you make your way up an ever growing tower by moving left and right and jumping on pads as you dodge lasers beams, collect tokens and more. It seems fairly simple but it's neat how each room is created by Nano Banana. You can scan the QR code pictured above if you want to give it a try for yourself.
Google I/O 2026 keynote countdown has begun! What do you expect?
Howdy folks! John Velasco here for Tom’s Guide and I’ll be taking you all through everything that’s going on with the Google I/O 2026 keynote. Last week’s Android Show certainly delivered a lot of buzz with announcements around Android Auto, Gemini Intelligence, Googlebooks, and more.
My colleagues Jason England and Paul Antill are on the ground to get first hands-on experiences with whatever else new that Google reveals at its keynote? I’ve been covering Google I/O since the very first one — and while it’s always been an Android first affair — we’ve seen how it’s expanded to other exciting areas.
'Take a second at how far we've come'
Google I/O 2026 kicks off with another inspirational video showing the impact of Google's products and services, with real world implications. It concludes saying, "let's make something that matters"
9.7 trillion tokens a month processed
Google's Sundar Pichai steps out onto the stage with opening remarks, including a funny clip of his highlights over the past year.
Pichai goes onto explain how Gemini is being access by more people than ever, with 9.7 trillion tokens a month.
Gemini app usage continues to grow at a rapid rate
The Gemini app usage continues to grow and the numbers shared by Sundar Pichai is proof of that, surpassing 900 million active users with the Gemini app, which is 2x growth over the past year.
Ask YouTube rolling out — makes video searches better
Google's rolling out a new feature with YouTube called "Ask YouTube," which intelligently finds videos with the help of AI. These are digestible clips that are to the point, based on your questions and queries.
Docs Live makes editing easier with AI
Docs Live is another powerful tool that makes creating documents easier with your voice using AI help. The demo looked pretty convincing at how well it functions.
TPUs make a huge impact to processing tokens
Most consumers might not think much about how tokens are processed, but the numbers are staggering. With more people leaning on AI, Google's investing in more TPU development to keep up pace, which as Sundar Pichai bluntly says, makes a huge difference with latency.
Gemini Omni is a new model that's multi-modal
Demis Hassabis is now on stage to talk more about models, with a focus on World models, which will be used to train AI bots. He goes on to talk about its latest model: Gemini Omni,
With Omni, its multi-modal capability can help generate videos using prompts, but it can be further refined through subsequent steps. You can edit videos with conversational languages, taking a simple selfie and transforming it to something else.
AI content verification is coming to Chrome
This new feature can help better tell you if content you see is generated with AI or not. OpenAI is now going to be using Google's content verifcation, in addition to other partners.
Gemini 3.5 Flash is even faster versus previous models
Gemini 3.5 Flash combines frontier intelligence with actions. This model has been tweaked to work much faster with its output speed, beating other models. It's 4x faster in fact.
Google Antigravity
Google Antigravity updates make programming easier for developers
Updates to Google Antigravity aims to assist programmers and developers with complex coding tasks. With this update, it's helping coders to create operating systems, including the ability to play Doom.
Gemini Spark handles your digital life
Gemini Spark is coming to the Gemini app, a persoanl agent that handles you digital life. It runs on virtual machines in Google Cloud and works 24/7. It's powered by Gemini 3.5.
Collaboration is going to get a big boost with Gemini Spark
The live demo of Gemini Spark shows how this agent works across different services, like Google Docs, Gmail, and more, all to populate details and information. One example Google shows is trying to organize a party with Gemini Spark, which populates an RSVP tracker created in Google Sheets, including real time changes to it. Spark can also send follow up reminders to people who haven't RSVPed.
Gemini Spark works across devices, iPhone and Android, in the background
Spark works on Android and iPhone, which syncs across all of your devices. You basically send it your ideas and it does the rest. The demo on stage shows how you can send voice actions and Spark creates tasks in the background for it to run when you're not using your devices.
New Google AI Ultra plans revealed
If you intend on using these new models, Google's introducing new Google AI Ultra plans starting at $100 per month.
Google Search is now backed by Gemini 3.5
Next up, Google Search! It's now being turbocharged with Gemini 3.5, with the ability to perform complex tasks through search, like creating an itinerary.
'We're entering the era of search agents'
"We're entering the era of search agents." Google Search is evolving by tapping into multiple search agents, so you can get more done. One of the examples shows how Google Search works with different agents in the background, like learning about stock price changes that happen in real time, along with information that's relevant to your searches.
'The best version of search is the one that works for you'
Robby Stein from Google is on stage to demo agentic capabilities of Gemini. "We believe the best version of search is the one that works for you," explains Stein.
With this new AI overview, you can get interactive experience in search, like learning about black holes. Search will even generate animations and custom UIs on different search topics.
Search can create tasks at scale
Search has the ability to help create mini apps for you, perfect for handling tasks at scale. The example of this is finding ideas for family fun things to do during the weekend, and Search creates a weekend planner from scratch all for you.
Universal Commerce Protocol open standard for shopping
Vidhya Srinvasan from Google is on stage to talk about shopping with Google, which all starts with Shopping Graph. Universal Commerce Protocol is an open standard that's being rolled out to several Google products.
'Universal Cart is a truly intelligent cart'
"Universal Cart is a truly intelligent cart," explains Srinvasan. This new feature will look up price changes, history, and adjust you price carts with help from agents. What's really fascinating is how intelligent it is, like know that the motherboard you're buying isn't compatible with processor you've added, so it'll automatically know that and provide you with choices.
Gemini app gets updated with Neural Expressive experience
The Gemini app is getting big changes, including a new experience built from the ground up with Neural Expressive. design language. It's a visual change that will include interactive images with its responses, timelines with generated videos, and more. Neural Expressive is rolling out to Android, iOS, and the web.
Gemini Omni is also coming to the app, like the ability to create and edit videos.
Gemini Omni's ability to generate video is impressive
Gemini Omni got demoed at I/O, showing how you can create a video you've uploaded with additional prompts to transform it to something else, like being able to rotate the framing and adding other elements.
Daily Brief finds important information
Daily Brief finds the most important things in you data, and presents them in digestible form.
Gemini app is coming to MacOS
If you're using MacOS, you'll be happy to know that the Gemini app is coming to it. Apple users will be able to access many of the new agents and features with Gemini on MacOS.
Google Pics and Stitch makes creation a breeze
Google Pics is an image creation and editing tool, from flyers to infographics. It's rolling out this summer.
Meanwhile, Stitch turns rough ideas to beautiful designs. For example, it can be used to create interfaces using prompts or your voice. You can create web sites in real time, using powerful tools to design a website.
Gemini Omni, Agent, Tools, and Music are rolling out to Google Flow
If you're looking to create more videos, Google Flow is getting a big boost to generate new characters and elements by simply typing your request. On top of that, your agents can take multiple actions at one - like taking a single image and transforming it into several different clips. You'll be able to edit at scale.
Google Flow Music takes your boring music and makes it professionally mixed
Another exciting demo shows Google Flow Music takes and how a simple piano track can be transformed into a complete musical track that sounds professionally mixed.
'Our first audio glasses will arrive this fall'
If you're eager to know about about Android XR, Google says that "our first audio glasses will arrive this fall." So farm Gentle Monster and Warby Parker are partners in creating these glasses in conjunction with Samsung. And yes, they'll pair with Android and iOS devices.
First look at Samsung's new 'intelligent eyewear'
Samsung's smart glasses introduces intelligent eyewear, with a focus on design and style. Really, the premise reminds us awfully of the Amazon Echo Frames - since it's an audio-focus experience.
Order coffee through your glasses with Gemini
Maps integrate with Samsung's intelligent eyewear, including getting directions. Gemini can also perform automated tasks, just like on the Galaxy S26, like using Doordash for a coffee order. The way it functions is basically the same as Gemini on the Galaxy S26, so it's not something special we haven't experienced before. Finally, it understands contextual information. But the biggest reveal is that it has a camera!
Computational Discovery helps to make deliver accurate predictions
Gemini for Science offers tools to accelerate research and make predictions, like hurricanes.
Will shopping habit change with Google's Universal Cart?
@tomsguide ♬ original sound - Tom’s Guide
Universal Cart could change the way we all shop online, which uses AI to proactively check your cart. Just take a peek at the demo of it on stage during Google I/O 2026 to see how the AI knows what parts you're buying for a new PC build.
Should Meta be afraid?
Google's collaboration with Samsung brought two working demos of its "intelligent eyewear" devices that aim to bring Gemini in a whole new way with glasses designed from Warby Parker and Gentle Monster.
What's surprising about them, is the deep integration of Gemini to perform tasks all through voice commands. What do you think? Are they enough to make Meta and its partners afraid?
Gemini Spark is the assistant I've been waiting for
Google Spark is one of the latest AI agents introduced during the keynote. What's special about this one is that it lives in the cloud, where the agent will proactively work on tasks on its own. It behaves differently because it'll continue to work in the background.
Personally, it's the AI agent I've been waiting for because it means that I can send it tasks and it'll track, manage, and proactively work on it. If you rewatch the live event, you'll see how it can be used to help with wedding plans, including the ability to track RSVPs, send reminders to people, and more.
YouTube's about to get even more helpful
@tomsguide ♬ original sound - Tom’s Guide
Just like everyone else, I spent a great deal of time on YouTube watching all sorts of videos. But a new feature coming soon will make it even more helpful.
Ask YouTube is an AI feature that better tailors results based on your search. For example, if you're wondering how to properly clean the sensor in a specific camera, you can send this inquiry to YouTube and it'll provide you with meaningful video results. Not only will it find videos that directly answer your question, but it'll even start the video at the exact place where it answers it.
Checking out AI mode
We are on site at Google I/O and checking out the various AI tools and devices that were announced today.
In this image, we're being shown the planner features in Google Search that can create weekend planners for you.
It's part of the new ability for Search to create mini apps for you and while connecting to other Google apps like your calendar.
Be sure to follow our TikTok
@tomsguide ♬ original sound - Tom’s Guide
Tom's Guide is on TikTok where we're showing off what we see at Google I/O including hands-on demonstrations with smart glasses, new Search tools and more.
Be sure to give us a follow if you're on that platform.
Eyes on with Google's prototype glasses
We've had a chance to go eyes on with a prototype of Google's smart glasses.
The display was turned off so we could experience it as a version of the future audio only model. It does seem reminiscent of the Ray-Ban Meta Wayfarer shades.
We didn't order any cold brews though.
Project Aura makes an appearance
Google and its partnerships have at least three smart glasses launching this year.
One of them is Xreal's Project Aura. You can see the display glasses above with a puck that is supposed to feature a Qualcomm Snapdragon chip that helps enable many of the features we can expect.
Live demo of Google's Audio Glasses
@tomsguide ♬ original sound - Tom’s Guide
During the Google I/O keynote speech, there was a live demo of Google's audio glasses showcasing the Gemini capabilities without the need to pull out your phone.
Google AI Ultra changes the game
During Google I/O 2026, the company unveiled a new AI Ultra tier that costs $100 per month. It's meant for devs, creators and power users.
This includes the newly unveiled Gemini Spark model which Google calls a "24/7 AI agent."
My colleague Amanda Caswell gets into way more depth here suffice to say, Google is committed to AI agents and assistants that actively manage information for users.
Intelligent eyewear and Gemini edit a photo
While testing out Intelligent Eyewear we gave it a prompt to turn my colleague Paul Antill.
The prompt was, "“Take a picture and make him look like a superhero."
It did, but it also changed the name on his media tag to read "Paul Arnold."
So meet Super Paul Arnold or GL, the Googler? Also, it made his face weird more cartoonish.
Hands-on with Google's 'Intelligent Eyewear'
The dust has settled on Google's I/O 2026 keynote, but the developer conference will continue to run for the next few days as we start to get our hands on the new tools revealed on stage. One of the most interesting is Google’s “Intelligent Eyewear,” and how it stacks up against the Ray-Ban Meta smart glasses.
Our Computing Editor, Jason England, is out at I/O this week and has got some hands-on time with the prototype and you can read his initial verdict right here. Bottom line? Mark Zuckerberg has a real fight on his hands.
Google's new Search is great — unless you're a journalist
We got to take a look at Google's redefined Search box at I/O and it's fair to say the added power of AI tools is an exciting prospect when it comes to making sense of a noisy and confusing internet.
But what's clearly a generational leap in making the wider internet more personable and accessible also has big implications. It could be a very worrying sign for publications and other businesses that exist on referral traffic from Google to survive. One thing's for sure: the days of seeing 10 blue links beneath your Google search are well and truly dead.
Could we see more improvements to the Android desktop mode?
Last year, Google quietly revealed its new Android desktop mode, which was developed developed with the help of Samsung. We could possibly hear more about what other changes and improvements is in store for the desktop mode during day 2 of Google I/O 2026.
Available to the most new Pixel phones, we've tried out this desktop mode and said it's a good start, but certainly trails what Samsung DeX offers. Since then, we've seen Motorola's version get updated in a big way with Smart Connect for the Razr Fold.
Meta has a lot of work ahead of itself
With its debut during the keynote yesterday, Google's positioning itself in a very good way with its upcoming Intelligent Eyewear devices, developed in collaboration with Samsung.
While Meta has been in the business of making smart glasses for a long while now, with notable models like the Ray-Ban Meta and Meta Ray-Ban Display, there are features they don't have that Google's upcoming glasses offer. Clearly there's a hump to overcome, but with Meta Connect 26 happening in September, it's going to need to catch up in a big way.
Googlebook was a no-show at Google I/O 2026
Google absolutely stirred some attention when it revealed its Googlebook during last week's Android Show, but it was a no-show at the Google I/O 2026 keynote. While these premium laptops are fundamentally tied to Gemini Intelligence, we were hopeful to learn more about them, especially what its hardware partners intend to do with their versions.
Xreal’s Android XR glasses could make you rethink about Google's intelligent eyewear
There's a lot of different types of smart glasses vying for consumer recognition and Android XR is one of them, which shouldn't be mistaken for Google's "intelligent eyewear" glasses it showed off at the Google I/O 2026 keynote.
Our very own Jason England got a first hands-on look at Xreal's Project Aura, which are AR glasses in the same vein as Samsung's Galaxy XR and Apple's Vision Pro headset. But unlike them, Xreal's version looks more like an ordinary pair of glasses that looks less conspicuous.
After trying them out at Google I/O 2026, England's quite impressed by them, so much so that he says that "these are a glimpse into the exciting future of AR glasses."
Gemini Omni takes video generation to the next level
If you love video as much as we do, then you might know how involved it can be putting together a compelling video that engages people. With Gemini Omni, it's taking video creation to a whole new level.
This generative service can do amazing things, like take a photo you give to it and generate several videos based on different prompts. You can also reimagine what they look like if you're looking for something entirely unique. Either way, check out what it's capable above in the video above.
Apple can't afford to lose more ground to Google
Google continues to push the boundaries for AI, evident in all of the new Gemini models and features introduced at the I/O 2026 keynote and Gemini Intelligence features from last week's Android Show.
Considering how Google Pixel phones have some of the best AI features around, it really puts the pressure on Apple, which is slated to host its WWDC 2026 developers conference next month. We know that Apple's going to tap Gemini to power its new Siri, but what to degree is still unknown. Nevertheless, Apple can't afford to lose more ground in the AI race.
Android Auto is about to be way more dynamic
CarPlay Ultra was meant to shake things up when it was first introduced last year, with an emphasis on deeper integration with cars, but its lack of support from car makers has fizzled it somewhat.
Android Auto, on the other hand, is about to get its biggest makeover in years with a new dynamic interface that adapts to different sized displays — and not to mention, multiple screens within a car
